The human papillomaviruses (HPV) are major etiologic agents of several human cancers and have been found in >95% of cervical cancers, >50% of other anogenital cancers, and 20% of oral, laryngeal, and nasal cancers. Cervical cancers are associated with a high risk subset of HPVs of which HPV-16 is the most common. The E6 and E7 viral oncogenes interfere with the functions of p53 and pRb, respectively, and thus disrupt regulation of the cell cycle and promote genomic instability. Progression of HPV-16 infection to high grade lesions and invasive cervical cancer requires significantly long lag periods, suggesting that multiple genetic events are required for malignant progression. The goals of this project are 1) to identify changes in cellular gene expression induced by HPV infection and 2) to identify changes in cellular and papillomaviral gene expression in cervical epithelia during progression of low grade squamous intraepithelial lesions (LGSIL) to high grade SIL (HGSIL) to invasive cervical cancer. Our ultimate goals are to identify biomarkers for diagnosis and prognosis as well as to identify pathways disrupted during cancer progression. We have pursued the first of these goals using cDNA microarray analysis of cellular gene expression in various tissue culture systems. Expression data thus generated were validated using Real Time Quantitative RT-PCR (QRT-PCR) for selected genes. Two immortalized cervical epithelial cell lines (W12 and C33A) were compared with primary foreskin keratinocytes. W12 cells are derived from a low grade cervical lesion and contain episomal HPV genomes. C33A cells are HPV negative cervical carcinoma cells that have mutated p52 and pRb genes. Many genes were differentially expressed in both W12 and C33A cells. These genes may represent common features of epithelial cell immortalization that could be induced either by viral infection or by mutation in cellular genes. Among these genes are known downstream targets in the p53 and pRb pathways. Another common feature is the down-regulation of interferon-induced genes, which may serve as an important mechanism for immune evasion. Interferon-induced genes also include anti-proliferative genes; down-regulation of these genes may be important for the continued growth of immortalized cells. A smaller number of genes were found to be differentially expressed only in HPV 16-positive W12 cells. These genes may represent specific targets of HPV infection and are involved in many pivotal cellular pathways such as cell cycle control, apoptosis, and transcription. We expect these studies to give considerable insight into cellular pathways that are targets of HPV proteins. Additional analyses of both acutely infected keratinocytes as well as keratinocytes stably maintaining HPV genomes are in progress. To carry out our second goal, we are currently using Laser Capture Microdissection (LCM) to obtain pure populations of normal, precancer, and invasive cervical cancer cells for expression analysis. RNA is then isolated and gene expression compared with a reference RNA pool using cDNA microarrays. We developed a method for signal amplification that takes small amounts of RNA from microdissected tissues and generates sufficient RNA for microarray assays. This method uses a combination of 10 cycles of PCR and in vitro transcription with T7 polymerase. Multiple pilot microarray assays were undertaken to compare the standard method using large amounts of total RNA without amplification with the amplification method using small amounts of sample RNA. Both methods gave similar gene expression profiles (correlation coefficients between 0.81 and 0.91), indicating that the limited amplification causes minimal distortion of the gene expression profile. In preliminary studies, 9 samples of invasive squamous cell carcinoma (SCC) of the cervix were obtained using LCM and analyzed using microarrays. Hierarchical clustering and multi-dimensional scaling analyses closely clustered all samples of cervical SCC and normal cultured cervical epithelial cells into two groups with highly distinguishable gene expression patterns. Although the invasive cervical SCC samples have similar gene expression profiles overall, there were some differences. More analyses will be required to determine if there are molecular subclasses of invasive cervical carcinoma. Future work will analyze many normal, precancer, and cancer tissues and attempt to identify molecular subclasses of these lesions and the biomarkers that distinguish them. This work will be done as part of a large epidemiological study in collaboration with the Division of Cancer Epidemiology and Genetics, NCI. Finally we are investigating changes in alternative splicing of selected cellular genes during cervical cancer progression. For these studies we have developed isoform-specific real time QRT-PCR assays for VEGF, hTERT (catalytic subunit of telomerase), and Bcl-X. The alternative isoforms of these proteins are expected to have significantly different biochemical properties. These assays will be applied to normal, precancer, and cancer tissues.